Tractian is now a unicorn. The Atlanta company, which sells sensors and AI that warn plants before their machines break, raised a $130 million Series D at a $1.4 billion valuation, announced on October 22. Sapphire Ventures led, with M&G’s Catalyst as co-lead. N47, General Catalyst, Bosch Ventures, 53 Stations, Monashees, Endeavor Catalyst, Norte Ventures and Y Combinator also put money in.
The valuation roughly doubled in under two years. Sapphire also led the $120 million Series C in December 2024, which valued Tractian at $720 million. The Information reports that annual recurring revenue passed $100 million in September. That is the number that turned a sensor startup into a billion-dollar company.
What Tractian actually sells
The hardware is plain. Wireless sensors bolt onto motors, pumps, compressors and conveyors and record vibration, ultrasound, temperature and electrical signals. The value is in what happens next. Tractian merges those readings into a single condition picture for each machine, and an AI layer trained on what the company calls one of the largest industrial failure datasets names the likely root cause and suggests the fix. The fix then becomes a work order in Tractian’s own maintenance management software, or in SAP, Maximo or Oracle NetSuite if the plant already runs those.
That last step matters more than it sounds. Plenty of companies sell vibration sensors. Few own the loop from “this bearing is going” to “a technician has the job, the part is in stock, and the downtime is scheduled”. Tractian sells sensors, diagnosis and maintenance software as one package, which makes it harder to swap out once a plant is wired.
Customers include Eli Lilly, McKesson, Whirlpool, Cummins, Embraer and Coca-Cola bottlers, along with data centers and government sites it doesn’t name. The company says more than 2,000 facilities use it. It has more than 700 employees, offices in Monterrey, São Paulo and Poland, and a new headquarters going up in Midtown Atlanta.
Why investors pay up for broken machines
Unplanned downtime is one of the most expensive problems in heavy industry. Siemens put the cost at about 11% of annual revenue for the world’s 500 largest industrial companies, roughly $1.4 trillion a year. Industrial equipment in service averages 24 years old. The people who used to hear a failing bearing before it seized are retiring, and Deloitte expects more than 2 million U.S. manufacturing jobs could go unfilled by 2030. An AI that does a veteran mechanic’s listening is an easy sale in that market.
Tractian also rides the “physical AI” label that investors attach to anything where models meet machines. That’s where most of the money in industrial tech is going this year. The difference here is revenue. Tractian sells something that pays for itself the first time it catches a failing pump at a pharma plant or a cooling unit at a data center, and buyers can measure it in hours of downtime avoided.
The Brazilian route to a U.S. unicorn
Leonardo Vieira, Igor Marinelli and Gabriel Lameirinhas founded Tractian in Brazil in 2019. Y Combinator backed it early, and Marinelli, now CEO, later moved the company’s base from the Bay Area to Atlanta. That path is rarer than it should be. Latin American founders usually build for their home market first. Tractian built in Brazil and Mexico, where factories are cost-sensitive and plant managers are hard to impress, then took the product to U.S. manufacturers that run older equipment on thinner maintenance crews.
The competition is crowded. Augury is its closest rival in AI machine health. Siemens, ABB, Emerson and Honeywell all sell condition monitoring, and plant software vendors keep adding predictive features. Tractian’s bet is that a single product covering sensor, diagnosis and work order beats a big vendor’s catalog of parts that don’t talk to each other.
The Series D money goes to scaling that platform across more industries and more plants. Pharma, food and beverage, oil and gas, energy, water, defense and data centers are all on the target list.
Old machines aren’t going anywhere. Somebody has to keep them running.